DE102007049660B4 - X-ray tomography device - Google Patents

Abstract

An x-ray tomography apparatus (10) comprising: scanning means (102, 103) for irradiating an object (HB) with X-rays while at least one of a gantry (100) and a table (109) moves along a body-axis direction of the subject (HB) in order to thereby generate projection data of the object (HB), a CT value change specifying means (26) for specifying a magnitude of change in the CT value in the body axis direction between a plurality of tomographic images obtained by back projection of the projection data Relating to each of the pixel areas included in the tomographic images; an image processing condition selecting means (28) for selecting an image processing condition for performing an image processing process for reducing artifacts in accordance with the specified amount of CT value change; and artifact reduction means for performing an image processing process using the image processing condition selected by the image processing condition selection means.

Description

BACKGROUND OF THE INVENTION

The present invention relates to an X-ray tomography device displaying tomographic images that are less susceptible to artifacts, such as artifacts. For example, cone beam artifacts, windmill artifacts, etc. that occur in an X-ray CT (computed tomography) apparatus or the like, as well as an associated artifact reduction method.

In a multi-slice x-ray computed tomography (x-ray CT) apparatus, the number of slices has now risen to 64 or 256. Heretofore, various cone beam image reconstruction algorithms are known in which a spiral scan performed by the X-ray CT apparatus is used. However, a problem common to all cone beam image reconstruction algorithms is that a sampling interval in a body axis direction (also referred to as a z direction or slice direction) of an object is not sufficient. These algorithms contradict the Nyquist theorem and cause vortex-like windmill or windmill-like artifacts within each reconstructed image due to high-frequency components. This means that interpolation calculations can not be performed in an ideal way if the resolution of a detector is insufficient for a structure and a spiral pitch factor (feed per revolution) is increased in the spiral scan so that windwheel-like artifacts appear in the image.

In order to reduce such pinwheel-like artifacts, multipoint z-directional interpolation is performed to reduce the fluctuation values of a target signal, thereby forming the pinwheel-like artifacts into a shadow or grayscale zone, respectively. In the EP 1 351 192 A1 is z. For example, after performing a reconstruction function convolution process, an interpolation process is performed in a z-direction to thereby try to reduce the pinwheel-like artifacts.

However, in the method of performing the multipoint interpolation in the z-direction for reducing artifacts, interpolation is performed even in image areas in which artifacts do not arise, resulting in a reduction in resolution in a z-direction Increasing the resolution no clear tomographic image can be obtained.

US 2006/0029285 A1 , which is considered as the closest prior art, describes a system and method for processing a medical image to reduce artifacts. Multiple first images are acquired, which together define a first image volume, and the first multiple images are low-pass filtered to obtain multiple second images. Subsequently, a first image of the plurality of first images and a corresponding second image of the plurality of second images are selected, and the second image is added with a noise lost in the low-pass filtering. Based on the pixel values in the plurality of second images, a gradient image is determined having a gradient value at each pixel location in the second image. Based on the gradient image, the first image and the associated second image are reconstructed with restored noise to obtain a corrected image that does not contain imaging artifacts.

It is an object of the present invention to provide an X-ray tomography device which safely removes artifacts without lowering the resolution in a z-direction, thereby reducing the artifacts.

BRIEF DESCRIPTION OF THE INVENTION

This object is achieved by the X-ray tomography device having the features of patent claim 1. Particularly preferred embodiments of the invention are the subject of the dependent claims.

In the present invention, artifacts are reduced only with respect to those image areas from a three-dimensional backprojected tomographic image in each of which the artifact has been generated. The three-dimensionally backprojected tomographic image is used as it is with respect to an artifact-free area, and the tomographic image is displayed. Therefore, with respect to the image area free of artifacts, a clear tomographic image can be obtained without reducing the resolution in a z-direction.

An X-ray tomography apparatus according to a first embodiment includes a scanning device for exposing an object to X-ray while moving at least one gantry and / or a table along a body-axis direction of the object to thereby generate projection data of the object, a CT value change specifying device for specifying a size of the change of the CT value in the body axis direction between a plurality of tomographic images obtained by backprojecting the projection data with respect to each of the pixel areas included in the tomographic images, an image processing condition selecting means for selecting an image processing condition to perform an image processing process for reducing artifacts according to the specified size of the CT value change and an artifact reduction means for performing an image processing process using the image processing processing condition selected by the image processing condition selection means.

In the X-ray tomography apparatus according to the first embodiment, the amount of change of the CT value in the body axis direction with respect to each of the pixel areas of the back-projected tomographic image is accurately determined. Based on the determined size of the CT value change, an image processing condition is selected. Based on the specified size of the CT value change, an image area with resulting artifacts and an image area without artifacts are determined. If no artifacts occur, an image processing condition is selected that uses an image in that image area as is. The image processing can be performed only on the image area with the artifacts generated therein, so that the artifacts are reduced. Incidentally, the amount of change of the CT value includes the mean value of the change of the CT value in the body axis direction or the difference between the maximum and the minimum value of the CT value change in the body axis direction.

The X-ray tomography apparatus according to a second embodiment further includes artifact ratio calculating means for calculating a ratio in which each pixel area in which the magnitude of the CT value change is in a predetermined range is occupied within a predetermined range included in each of the tomographic images and determining means for determining that, when the ratio is greater than a threshold value, an image processing process for reducing the artifacts at the pixel area is performed.

Although in the X-ray tomography apparatus according to the second embodiment, there is a high probability that when the size of the CT value change is within the predetermined range, there will be pixels with artifacts, the magnitude of the CT value change may be increased due to an influence, e.g. An imaging condition, fall within the predetermined range. Thus, it is calculated in which ratio a pixel area in which the amount of change of the CT value is in a predetermined range occupies a predetermined area included in each of the tomographic images. When the ratio is low, a high-resolution pixel viewed in a body-axis direction as it is is maintained without performing an artifact reduction process. If the ratio is high, there is a high probability that artifacts will be generated. Therefore, it is possible to apply the image processing only to an image area of its tomographic image in which artifacts are generated to reduce the artifacts.

In the X-ray tomography apparatus according to a third embodiment, the image processing process of the artifact reduction means includes a process of multiplying the image areas of the plurality of tomography images by weighting factors respectively selected as the image processing condition and adding the multiplication results.

In the embodiment according to the third embodiment, it is determined that when the amount of change of the CT value is in a predetermined range, there is an image area having artifacts generated therein. With respect to each image area in which artifacts are generated, multiple pixel areas in a body axis direction are multiplied by weighting factors, and the results of the multiplications are added together to thereby reduce the artifacts in the pixel area of each tomographic image.

In the X-ray tomography apparatus according to a fourth embodiment, the image processing condition selection means sets the weighting factor to be greater than zero when the size of the CT value change is within the predetermined range.

In the embodiment according to the fourth embodiment, as for the pixel area having artefacts generated therein, a plurality of pixel areas in a body axis direction with weighting factors multiplied greater than zero, thereby reducing artifacts in the pixel area of each tomographic image.

According to a fifth embodiment, the image processing condition selection means changes the weighting factor according to the number of pixel areas in the body axis direction.

In the embodiment according to the fifth embodiment, the weighting factor may be changed on the basis of the number of pixel areas in the body axis direction, for example, in the case of a single layer image in the vicinity of a target area 3 and in the case of n layer images in the vicinity thereof 2n + 1 equivalent.

In the X-ray tomography apparatus according to a sixth embodiment, the image processing condition selection means changes the weighting factor according to the size of the CT value change according to the third or fourth embodiment.

In the embodiment according to the sixth embodiment, the weighting factor may be determined according to the magnitude of the change of the CT value, e.g. For example, according to the amount of change of, for example, 20 HU (Hounsfield units), 100 HU, or the like. This means that suitable artifact reductions can be achieved in each case in an image area in which artifacts are generated to a great extent, as well as in an image area in which artifacts hardly occur.

In the X-ray tomography apparatus according to a seventh embodiment, the CT value change specifying means determines an index value corresponding to the amount of change of the CT value.

In the embodiment according to the seventh embodiment, an index value is determined based on the size of the CT value change. Thus, the index value allows discrimination between an image area having artifacts generated therein and an image area without artifacts. Using the index value makes it easier to set the image processing.

In the X-ray tomography apparatus according to an eighth embodiment, the plural pixel areas in the body axis direction are multiplied by weighting factors related to the pixel areas corresponding to the index value, and the multiplication results are added up.

In the embodiment according to the eighth embodiment, the weighting factors are changed with respect to the plural pixel areas in the body axis direction according to the index value, and they are multiplied therewith, thereby making it possible to reduce artifacts in the pixel area of each tomographic image. This means that suitable artifact reductions can be applied in each case to an image area in which artifacts are generated to a great extent, or an image area in which artifacts are scarcely generated.

According to a ninth embodiment, when the magnitude of the change of the CT value in the second or fourth embodiment is in the range of 3 HU to 300 HU, the artifact reduction means reduces artifacts with respect to each pixel area.

It is judged that when the magnitude of the CT value change is in the range between 3 HU and 300 HU, the artifacts are generated. For example, assuming that the magnitude of the change in the CT value is 300 HU or more, it means a region with a change from a soft tissue to a bone or the like, or an inverse change region. If the amount of change in the CT value is 3 HU or less, it means that the soft tissue or the bone is continuous in multiple slice directions. On the other hand, changing the CT value in the range of 3 HU to 300 HU means that a windmill-like artifact or a cone beam artifact is generated.

In the X-ray tomography apparatus according to a tenth embodiment, each pixel area has a single pixel or a plurality of pixels.

In the tenth embodiment, a pixel area of a target tomographic image may be processed in the form of a single pixel or a plurality of pixels.

A method for reducing artefacts of X-ray-generated tomographic images according to an eleventh embodiment comprises the steps of obtaining a magnitude of change in CT value in a body axis direction between a plurality of tomographic images obtained by backprojection of projection data of an object obtained by irradiation of the An object is acquired with X-rays while at least either a gantry and / or a table is moved along the body-axis direction of the object is specified with respect to each of the pixel areas contained in the tomographic images, an image processing condition for performing an image processing process for reducing the artifacts is selected according to the specified size of the CT value change and an image processing process is performed using the selected image processing condition.

In the artifact reduction method according to the eleventh embodiment, the amount of change of the CT value in the body axis direction with respect to each of the pixel areas of the back-projected tomographic images is accurately determined. The image processing condition is selected based on the size of the CT value change. That is, based on the specified amount of CT value change, a decision is made as to whether artefacts have been generated in an image area or whether an image area is free of artifacts. When it is determined that the artifacts have been generated, an image processing condition is selected that uses the image as it is. It is possible to apply the image processing only to the image area where the artifacts are generated, so that the artifacts are reduced.

The artifact reduction method according to a twelfth embodiment further comprises the steps of calculating how each pixel area in which the amount of change of the CT value is in a predetermined range is occupied within a predetermined range in each of the tomographic images and it is determined that if the ratio is larger than a threshold, the image processing process for reducing the artifacts is performed on each of the pixel areas.

Although there is a high probability that, in the case where the size of the CT value change is in a predetermined range, there are pixels with artifacts, the amount of change of the CT value may be varied depending on an influence such as an imaging condition. fall within the predetermined range. Thus, the artifact reduction method according to the twelfth embodiment calculates the proportion that the pixel area in which the amount of change of the CT value lies in the predetermined range occupies within the predetermined range included in each of the tomographic images. If the ratio is low, a high-resolution pixel as viewed in the body-axis direction is maintained as it is without performing an artifact reduction process. If the proportion is high, there is a high probability that artifacts will be generated. Therefore, it is possible to perform image processing only in an image area of a tomographic image in which the artifacts occur to reduce the artifacts.

In the artifact reduction method according to a thirteenth embodiment, the image processing process for reducing the artifacts includes a process for multiplying the image areas of the tomographic images by weighting factors respectively selected as the image processing condition and adding the results of the multiplications.

In the artifact reduction method according to the thirteenth embodiment, it is determined that when the amount of change of the CT value is within a predetermined range, there is an image area in which artifacts are generated. With respect to each image area in which artifacts are generated, plural pixel areas in a body axis direction are multiplied by weighting factors, and the multiplication results are added together, thereby reducing the artifacts in the pixel area of each tomographic image.

In the artifact reduction method for tomographic images formed by X-rays according to a fourteenth embodiment, the image processing condition selecting step sets the weighting factor to be larger than 0 when the amount of change of the CT value is within the predetermined range.

In the artifact reduction method according to the fourteenth embodiment, as for the image area in which artifacts are generated, a plurality of pixel areas in a body axis direction are involved Multiplied by weighting factors set greater than 0, thereby reducing the artifacts in the pixel area of each tomographic image.

In the artifact reduction method according to a fifteenth embodiment, the weighting factor is changed according to the number of pixel areas in the body axis direction.

In the artifact reduction method according to the fifteenth embodiment, the weighting factor may be changed based on the number of pixel areas in the body axis direction, for example, in the case of a single layer in the vicinity of a target area 3 and in the case of n layers in the vicinity thereof 2n + 1 equivalent.

In the artifact reduction method according to a sixteenth embodiment, the weighting factor is changed according to the amount of change of the CT value.

In the embodiment according to the sixteenth embodiment, the weighting factor may be set according to the size of the CT value change, e.g. As the change size of, for example, 40 HU, 60 HU or the like, to be changed. This means that suitable artifact reductions can be achieved in each case in an image area in which artifacts are generated to a great extent, or in an image area in which artifacts are scarcely generated.

In the artifact reduction method according to a seventeenth embodiment, an index value corresponding to the specified amount of change of the CT value is determined.

In the artifact reduction method according to the seventeenth embodiment, an index value is determined from the amount of change in the CT value. Thus, the index value is able to distinguish between an image area where artifacts are generated and an image area without artifacts. Using the index value, it is easy to specify image processing.

In the artifact reduction method according to an eighteenth embodiment, the weighting factor is changed corresponding to each region of the object.

In the artifact reduction method according to the eighteenth embodiment, the weighting factor may be changed depending on the head, the neck, the breast, and the like, even if the same amount of CT value change is detected. This is because the nature of the soft tissue, the thickness of the bone, the complexity of the shape and the like vary depending on the regions of an object.

In the artifact reduction method according to a nineteenth embodiment, when the magnitude of the change of the CT value is in the range of 3 HU to 300 HU, an artifact is reduced with respect to each of the pixel areas.

In the embodiment according to the nineteenth embodiment, it is judged that when the magnitude of the CT value change is in the range between 3 HU and 300 HU, artifacts have been generated. For example, if the size of the CT value change is 300 HU or more, this indicates a portion or region with a change from a soft tissue to the bone or the like, or a region with a reverse change. If the size of the CT change in value is 3 HU or less or 10 HU or less depending on imaging conditions, it means that the soft tissue or bone is continuous in multiple slice directions. In a pixel region with a transition from the soft tissue to the bone or vice versa from the bone to the soft tissue or a pixel region with a transition from the lung to a soft tissue, or vice versa, the magnitude of the change is in the range of over 200 HU to over 300 HU in front. Therefore, on the other hand, it can be estimated from the change in the CT value from 3 HU to 300 HU that a windmill-like or windmill-like artifact or a cone-beam artifact has been generated.

In the artifact reduction method according to a twentieth embodiment, in the case where the amount of change of the CT value is in the range of 0 HU to 300 HU, an artifact with respect to each of the pixel areas is reduced and noise with respect to the pixel area is reduced.

In the embodiment according to the twentieth embodiment, noise reduction is performed when the amount of change of the CT value is 3 HU or less. This means that in the case that the size of the change of the CT value is 3 HU or less, a soft tissue or a Bone continuously runs in several layers. However, if the amount of change of the CT value is 3 HU or less, weighting is performed on a plurality of slice images, so that noise reduction of each tomographic image is also performed.

With an X-ray tomography apparatus according to the present invention and an artifact reduction method, artifacts are reduced only with respect to image areas of a three-dimensional backprojected tomographic image each having the artifact generated therein. A tomographic image may be displayed in which the three-dimensionally back-projected tomographic image is used as it is with respect to an area free of the artefacts that occur.

BRIEF DESCRIPTION OF THE DRAWINGS

1 FIG. 12 is a block diagram showing a configuration of an X-ray CT apparatus. FIG 10 illustrated in accordance with the present embodiment.

2 Fig. 12 is a diagram showing geometrical arrangements showing an X-ray tube 10 and a multi-row X-ray detector 103 illustrate.

3 Fig. 12 is a flowchart schematically showing the process of taking a tomographic image in an X-ray CT apparatus 10 of the present invention.

4 shows a flow chart for the reduction of artifacts after the determination of backprojection data D3.

5 Fig. 11 is a conceptual diagram illustrating pixels of tomographic images based on back projection data D3 (x, y, z) and their pixel areas.

6 FIG. 15 shows an example in which a tomographic image D3 (x, y, z) before performing an artifact reduction process and a tomographic image D31 (x, y, z) subjected to the artifact reduction process are displayed 60 are displayed.

7 shows a diagram in which index functions are shown.

8th shows a diagram illustrating further index functions.

9 shows a schematic illustration showing tomographic images of the head of an object before they have been subjected to an artifact reduction process, and reconstruction areas P.

10 11 is a schematic diagram illustrating a flowchart for performing an artifact reduction process after artifact conditions.

11 shows a cross-sectional view in a body axis direction of the chest of an object HB up to his head and artifact ratios.

DETAILED DESCRIPTION OF THE INVENTION

<Configuration of the X-ray tomography device>

1 FIG. 12 is a block diagram showing the configuration of an X-ray computed tomography apparatus (an X-ray CT apparatus). FIG. 10 illustrated in accordance with the present embodiment. The X-ray tomography device 10 is with a gantry 100 and a table 109 equipped to serve an object or a subject HB in an imaging area of the gantry 100 contribute. The table 109 is moved in a z-direction which corresponds to the direction of a body axis of the object HB. The gantry 100 has a rotating ring 102 on and includes an x-ray tube 101 for irradiation with an X-ray beam XR leading to the orbiting ring 102 has the shape of a cone beam, and a multi-row X-ray detector 103 , opposite the X-ray tube 101 is arranged. The multi-row X-ray detector 103 detects X-rays transmitted through the object HB.

The multi-row X-ray detector 103 has scintillators and photodiodes. A data acquisition circuit 104 , which is generally referred to as DAS (data acquisition system), is with the multi-row X-ray detector 103 connected. For each channel in the data acquisition circuit 104 are an I-V converter for converting a current signal for each channel of the multi-row X-ray detector 103 to a voltage, an integrator for periodically integrating the voltage signal in synchronism with an X-ray irradiation cycle or period, a preamplifier for amplifying a signal output from the integrator, and an analog-to-digital converter for converting a signal output from the preamplifier is provided in a digital signal. Digital signals coming from the data acquisition circuit 104 are transmitted by a data transmission device 105 to an image processor 20 transfer.

On the side of the control panel is a high voltage generator 51 present, which provides the voltage for the X-rays. The high voltage generator 51 periodically generates a high voltage and supplies the high voltage via a slip ring 113 to the x-ray tube 101 ,

A scanning control device 53 on the side of the control panel performs several scan schemes, such as. As an axial scan, a spiral scan or a spiral scan with variable pitch factor, from. The axial scan is a scanning method in which the X-ray tube 101 and the multi-row X-ray detector 103 every time the table 109 is rotated by a predetermined pitch in the Z-axis direction, to thereby capture or acquire projection data. The spiral scan is a scanning method with movement of the table 109 at a predetermined rate in a state where the x-ray tube 101 and the multi-row X-ray detector 103 rotated to capture raw data. The variable-pitch spiral scanner is a scanning method with variation of the speed of the table 109 while the x-ray tube 101 and the multi-row X-ray detector 103 through a rotation mechanism 111 in a manner similar to the spiral scan to capture raw data. The scan control 53 controls the rotation mechanism 111 synchronous with the high voltage generator 51 on and performs a line on the scanning operations such. B. the periodic acquisition of raw data by the data acquisition circuit 104 etc, out.

An input device 55 includes a keyboard or a mouse through which inputs can be received by the operator. A storage device 59 stores programs, X-ray detector data, projection data and X-ray tomography images. The image processing device 20 leads to the projection data coming from the data acquisition circuit 104 are sent, a pre-processing process, an image reconstruction process, a post-processing process, and the like. A display or a display 60 displays a work screen and a picture reconstructed tomographic image.

<Configuring the image processing group>

The image processing group or device 20 contains a preprocessor 21 a beam hardening processor 23 , a three-dimensional rear projection processor 24 , an artifact reduction unit 25 , a CT value change specifying unit 26 , an artifact ratio calculating means 27 an image processing condition selecting means 28 and a determination device 29 ,

The preprocessor 21 corrects the inconsistent channel-to-channel sensitivity with respect to the raw data generated by the data acquisition circuit 104 and performs a preprocessing process such as B. X-ray dose correction, to correct for an extreme reduction in signal strength or a signal omission due to a strong X-ray absorber, mainly a metal portion. Incidentally, data that has been subjected to a preprocessing process is referred to as projection data in the present embodiment.

The beam hardening processor 23 causes a correction processing with respect to the beam hardening of the projection data. The beam hardening is characterized by the phenomenon that the absorption of the X-rays changes due to a penetration or transmission thickness, even when it is the same material, thereby the CT value (the luminance) on each CT image varied. This means in particular that the energy distribution of the radiation transmitted through an object is shifted towards the high energy side. Therefore, the beam hardening is corrected in a layer direction and a channel direction of the projection data.

The three-dimensional rear projection processor 24 receives the projection data from the preprocessor 21 preprocessed, and reconstructs images based on the projection data. The Projection data is subjected to a fast Fourier transform (FFT) to be converted into a frequency domain and convolved with a reconstruction function kernel (j), whereupon it undergoes an inverse Fourier transform. The three-dimensional rear projection processor 24 effects a three-dimensional back projection process on the projection data subjected to the convolution processing with the reconstruction function kernel (j) to determine a tomographic image (xy plane) for each body axis direction (Z direction) of the object HB. The three-dimensional rear projection processor 24 allows the storage device 59 to save the tomographic image.

The artifact reduction unit 25 reads the tomographic image after the three-dimensional backprojection from the storage device 59 and subject it to an artifact reduction process. The artifact reduction unit 25 allows the storage device 59 to save the artifact-reduced tomographic image and causes it to appear on the display 60 is shown.

The CT value change specification unit 26 specifies the amount of change in the CT value in the body axis direction. This is because, in the case that the magnitude of the CT value change in the body axis direction falls within a predetermined range, it can be estimated that artifacts are generated.

The artifact ratio calculator 27 calculates what proportion of an image area with artefacts generated in the tomographic image or an object area of the tomographic image.

The image processing condition selection means 28 selects which image processing condition is established when the artifact reduction unit 25 performs image processing to reduce the artifacts. The image processing condition selection means 28 For example, selects weights for assigning weights to multiple pixels as viewed in the body axis direction.

The determining device 29 determined from the ratio that by the artifact ratio calculating means 27 is calculated, whether the artifact reduction unit 25 should perform the artifact reduction process.

2 (a) and 2 B) show graphs showing the geometrical arrangements of the x-ray tube 101 and the multi-row X-ray detector 103 illustrate. 2 (a) is a diagram showing the geometrical arrangements of the x-ray tube 101 and the multi-row X-ray detector 103 shows how they appear from an xy plane, and 2 B) is a diagram showing the geometrical arrangements of the x-ray tube 101 and the multi-row X-ray detector 103 shows how they appear from a yz plane. An anode of the X-ray tube 101 generates an X-ray XR, called a cone beam. If the direction of a central axis of the cone beam is parallel to a y direction, this is assumed to be a view angle of 0 °. The multi-row X-ray detector 103 has X-ray detector rows corresponding to J rows in the z-axis direction (layer direction), e.g. B. 256 rows. Each of the X-ray detector rows has X-ray detector channels, the I channels, viewed in the channel direction, z. B. 1024 channels correspond. In 2 (a) More X-rays in the X-ray beam XR are emitted from the X-ray focus of the X-ray tube 101 is emitted at the center of an image reconstruction area P by a beamforming X-ray filter 121 while less X-rays are applied in the X-ray beam XR at portions around the image reconstruction area P. Thus, after spatially controlling the X-ray dose, the X-rays are absorbed in the object HB located within the image reconstruction area P, the transmitted X-rays from the multi-row X-ray detector 103 recorded as raw data.

In 2 B) XR XR is emitted from the anode of the X-ray tube 101 is emitted in the direction of the layer thickness of a tomographic image by an X-ray collimator 123 controlled, and so the X-rays are absorbed by an object HB, which is located in the vicinity of the central axis of rotation IC, wherein the penetrating X-rays through the multi-row X-ray detector 103 recorded as raw data. All raw data collected by the multi-row X-ray detector 103 are detected after the object HB has been exposed to X-rays are detected by the data acquisition circuit 104 A / D converted and then viewed from the multi-row X-ray detector via the data transmission device 105 the image processor 20 fed. The raw data included in the image processor 20 are fed from the image processor 20 in accordance with the corresponding program of the storage device 59 processed and reconstructed into a tomographic image, which is shown below on the display 60 is shown. By the way, can also be a two-dimensional X-ray detector having a matrix structure as represented by a flat-panel X-ray detector, although in the present embodiment, the multi-row X-ray detector 103 is used.

<Function chart for tomogram recording>

3 FIG. 12 is a flowchart showing the outline of the process of taking a tomographic image in the X-ray CT apparatus. FIG 10 illustrated in accordance with the present invention.

In step S11, a spiral scan is performed to the X-ray tube 101 and the multi-row X-ray detector 103 to rotate around the object HB and data from the multi-row X-ray detector 103 to capture while the table 109 is moved linearly. A z-direction position Ztable (view) is added to the raw data D (view, j, i) (where j = 1 to ROW and i = 1 to CH), depending on a view angle view, a detector line number j and of a channel number i, and the data acquisition is performed in a constant speed area.

In step S12, the raw data D0 (view, j, i) is subjected to a preprocessing process and converted into projection data. An offset correction, a logarithmic conversion, an x-ray dose correction and a sensitivity correction are performed.

In step S13, beam hardening correction is effected on the preprocessed projection data D01 (view, j, i) to thereby convert the data into projection data D1 which has undergone the beam hardening correction. The beam hardening correction in step S13 may be e.g. B. be performed by a multiplication with a polynomial. At this time, at each of the j rows, as in the slice direction of the multi-row X-ray detector 103 If an independent beam hardening correction can be performed, it is possible to correct the differences between the detectors placed in each row with respect to the X-ray energy characteristic when the X-ray tube voltages are different depending on the imaging conditions.

In step S14, on the projection data D1 subjected to the beam hardening correction, a z-filter folding process for applying filters in the slice direction (z direction) is performed, wherein the projection data D1 is converted into projection data D11 which undergoes the filter folding process to have. That is, the projection data of the multi-row X-ray detector 103 be subjected to a z-filter convolution process at each view angle and in each data acquisition system. When changing row-direction filter coefficients for each channel, layer thicknesses can be controlled depending on the distance to the image reconstruction center.

In step S15, convolution is performed on a reconstruction function kernel (j) with respect to the projection data D11 that has undergone the filter convolution process. That is, the fast Fourier transform (FFT) for converting the projection data D11 which has passed through the filter convolution process into a frequency domain, and the reconstruction function kernel (j) is applied to the projection data D11 by convolution. Then, the inverse Fourier transform is performed to convert the data into projection data D2 (view, j, i) which has been subjected to the reconstruction function convolution process. Because the convolution process for the reconstruction function kernel (j) and the reconstruction functions for each of the j rows of the multi-row x-ray detector 103 can be performed independently of each other, the differences between the noise characteristics and resolution characteristics in all lines can be corrected.

In step S16, a three-dimensional backprojection process is applied to the projection data D2 (view, j, i) having undergone the reconstruction function convolution process to determine backprojection data D3 (x, y, z). An image to be reconstructed becomes three-dimensional on a plane, i. H. an xy plane that is orthogonal to the z axis, reconstructed. The following reconstruction area P is assumed to be parallel to the xy plane.

In step S17, postprocessing processes such as artifact reduction process, CT value conversion and the like are performed on the backprojection data D3 (x, y, z) to obtain a tomographic image D31 (x, y, z). In the artifact reduction process, an image area in which artifacts are generated is detected by the amount of change of the CT value as viewed in the Z direction. A filtering process is performed only in the captured image area in which the artifacts are generated. With respect to an image area free of artifacts, the rear projection data D3 (x, y, z) is used as the tomographic image D31 (x, y, z) as they are.

<Artifact processing function chart>

4 FIG. 12 is a flow chart used to perform artifact reduction after determination of the backprojection data D3 (x, y, z). 5 Fig. 11 is a conceptual diagram showing pixels of tomographic images based on the rear projection data D3 (x, y, z) and their pixel areas. Incidentally, windmill or windmill type artifacts or cone beam artifacts can be reduced by the present flowchart.

In 4 (a) For example, a z-position of an object HB that an operator wants to confirm is specified in step S171. The artifact reduction unit 25 determines each pixel p (x, y, z) to be processed. If z. For example, assume that a square area of 512x512 pixels that is parallel to the xy plane represents a reconstruction area P as shown in FIG 5 (a) is shown, then x has a value range of 1 to 512, and y also has a value range of 1 to 512.

In step S172, the CT value change specification unit measures 26 the change of the CT value which is checked for each pixel p (x, y, z) to be processed in the z-direction. It is z. For example, suppose that the changes of the CT value in the z direction in the vicinity of a pixel to be processed p (x1, y1, z1) in the reconstruction area P are as follows: p (x1, y1, z-1) = 10 HU (Hounsfield unit, Hounsfield units) p (x1, y1, z) = 30 HU p (x1, y1, z + 1) = 50 HU

It can be seen that the amount of change of 40 HU results from the difference between the minimum CT value and the maximum CT value in the vicinity of the pixel p (x1, y1, z) viewed in the body axis direction.

Here shows 5 (a) the z-direction pixels in the vicinity of the pixel p (x1, y1, z1). The change in the CT value is explained below assuming a change for each pixel. However, an average CT value may be used in a pixel area (X1, Y1, Z1) in which several pixels surrounding a particular pixel are combined, or the highest CT or CT may be used. Value to be used. A pixel area consisting of several pixels is moved, being shifted for each specific pixel. Although the change amount of the CT value of a single layer in the vicinity of the pixel to be processed p (x1, y1, z1) is measured in the above, the change amounts of the CT values of n layers can be measured in the vicinity thereof.

Next, the CT value change specification unit determines 26 in step S173 an index. This index can be determined using the following function. In the following equation, the changes of the CT values of the n layers in the vicinity of the pixel to be processed p (x, y, z) are measured, and the intended index is determined from the changes. Index = f (p (x, y, z - n), p (x, y, z - n + 1) ... p (x, y, z) ... p (x, y, z + n )

That is, the index is set to reduce artifacts relative to the pixels in which the artifacts occur, while simultaneously setting the index to advantageously determine the pixel p (x, y, z) to be processed, as it is, in terms of using the pixels in which no artifacts occur. Functions for determining the indices are given below with reference to 7 explained.

It is assumed that if the change of the CT value as p (x1, y1, z-1) = 10 HU, p (x1, y1, z) = 30 HU and p (x1, y1, z + 1) = 50 HU, as in the previous example, the index = 1 is reached.

wobei g(i, Index) ein auf dem Index basierender Gewichtungskoeffizient oder -faktor einer i-ten Schicht in der z-Richtung ist. Beispielsweise setzt die Bildverarbeitungsbedingungs-Auswahleinrichtung 28 die Gewichtungsfaktoren für jede Schicht in der Umgebung des zu verarbeitenden Pixels p(x1, y1, z1) folgendermaßen fest:
Es wird angenommen, dass, wenn Index = 1 ist, der Gewichtungsfaktor g, der auf p(x1, y1, z – 1) angewendet oder diesem zugeordnet wird, g = 0,33 ist, der Gewichtungsfaktor g, der p(x1, y1, z) zugeordnet wird, g = 0,33 ist, und der Gewichtungsfaktor g, der p(x1, y1, z + 1) zugeordnet wird, g = 0,33 ist. Das heißt, dass jedes Pixel, in dem Artefakte entstehen, zu einem Pixel korrigiert wird, bei dem die in dessen Umgebung liegenden Schichtbilder gemittelt werden. Wenn n Schichtbilder verwendet werden, kann ein Wert von g = 1/(2n + 1) verwendet werden.Next comes the artifact reduction unit 25 image processing in step S174 on the pixel to be processed p (x, y, z) on the basis of the index value to determine a pixel p '(x, y, z) as a result of its processing. For example, the pixel p 'is expressed in the following Equation 1: [Equation 1]

where g (i, index) is an index-based weighting coefficient or factor of an i-th layer in the z-direction. For example, the image processing condition selecting means sets 28 the weighting factors for each layer in the vicinity of the pixel to be processed p (x1, y1, z1) as follows:
It is assumed that when Index = 1, the weighting factor g applied to or assigned to p (x1, y1, z-1) is g = 0.33, the weighting factor g is p (x1, y1, z), g = 0.33, and the weighting factor g associated with p (x1, y1, z + 1) is g = 0.33. That is, each pixel in which artifacts arise is corrected to a pixel in which the layer images in its environment are averaged. If n slice images are used, a value of g = 1 / (2n + 1) can be used.

It is assumed that when the index = 0.5, the weighting factor g associated with p (x1, y1, z-1) is g = 0.2, the weighting factor g, which is p (x1, y1 , z), g = 0.6, and the weighting factor g associated with p (x1, y1, z + 1) is g = 0.2. The influence of the pixel to be processed p (x, y, z) has a strong effect in every pixel in which a weak artifact occurs, but layer images in the vicinity of this pixel are also added to a small extent.

It is assumed that when the index = 0, the weighting factor g associated with p (x1, y1, z-1) is g = 0, the weighting factor g associated with p (x1, y1, z) g = 1, and the weighting factor g associated with p (x1, y1, z + 1) is g = 0. The pixel to be processed p (x, y, z) is set equal to each pixel without artifact, so that it is used as it is.

Incidentally, the weighting factors g (i, index) may be stored in a look-up table or the like or as predetermined functions based on information obtained by experiments or the like. be won. Since the previous or succeeding layer does not exist, the weighting factor can not be established in the manner described above on the initial layer or first layer or the last layer. Therefore, there is no need to perform a correction process on the first layer or the last layer, so that only each image in a lateral direction alone is used, and also to change the weighting factor.

In step S175, after the artifact reduction or reduction process, a tomographic image D31 (x, y, z) based on p '(x, y, z) is obtained. Then this will be on display 60 displayed.

6 FIG. 15 shows an example in which a tomographic image D3 (x, y, z) before performing the artifact reduction process according to the present embodiment and a tomographic image D31 (x, y, z) having undergone the artifact reduction process on the display 60 are displayed. The windmill type artifact and the cone beam artifact are clearly displayed on the tomographic image D3 (x, y, z). However, the influences of the wind type artifact and the cone beam artifact are reduced in the case of the tomographic image D31 (x, y, z) shown in the right drawing. In the tomographic image D31 (x, y, z) illustrated in the right drawing, a pixel area without artifacts becomes the same image as in the tomographic image D3 (x, y, z) shown in the left drawing so that it retains the same resolution.

The flowchart used in 4 (b) 2 is a flowchart in which no index or index function is used (see FIG 7 or 8th ), as described in step S173 of 4 (a) is described.

In the flowchart that is in 4 (b) Further, after the amount of change of the CT value, as viewed in the z-direction, is measured for each pixel to be processed p (x, y, z) in step S172, a weighting factor gv on the basis of the size is further illustrated the CT value change is determined in step S174 'without determining the index at the next location.

wobei gv(i, CTv) ein Gewichtungskoeffizient oder -faktor einer i-ten Schicht in der z-Richtung ist, der auf der Größe der CT-Wertveränderung beruht. Beispielsweise setzt die Bildverarbeitungsbedingungs-Auswahleinrichtung 28 die Gewichtungsfaktoren für jede Schicht in der Umgebung des zu verarbeitenden Pixels p(x1, y1, z1) wie folgt fest:
Es wird angenommen, dass, wenn die Veränderungsgröße des CT-Werts 40 HU beträgt, der Gewichtungsfaktor gv, der auf p(x1, y1, z – 1) angewendet oder diesem zugeordnet wird, gv = 0,33 ist, der Gewichtungsfaktor gv, der p(x1, y1, z) zugeordnet wird, gv = 0,33 ist, und der Gewichtungsfaktor gv, der p(x1, y1, z + 1) zugeordnet wird, gv = 0,33 ist.In step S174 ', the pixel to be processed p (x, y, z) is processed on the basis of the index value to determine a pixel p' (x, y, z) as a result of its processing. The pixel p '(x, y, z) is z. Expressed in the following equation 2: [Equation 2]

where gv (i, CTv) is a weighting coefficient or factor of an i-th layer in the z-direction based on the magnitude of the CT value change. For example, the image processing condition selecting means sets 28 the weighting factors for each layer in the vicinity of the pixel to be processed p (x1, y1, z1) as follows:
It is assumed that when the amount of change of the CT value is 40 HU, the weighting factor gv applied to or assigned to p (x1, y1, z-1) is gv = 0.33, the weighting factor gv, where p (x1, y1, z) is assigned, gv = 0.33, and the weighting factor gv associated with p (x1, y1, z + 1) is gv = 0.33.

It is assumed that when the amount of change of the CT value is 120 HU, the weighting factor gv assigned to p (x1, y1, z-1) is gv = 0.2, the weighting factor gv, which is p (x1 , y1, z), gv = 0.6, and the weighting factor gv associated with p (x1, y1, z + 1) is gv = 0.2.

It is assumed that when the magnitude of the CT value change is 200 HU, the weighting factor gv associated with p (x1, y1, z-1) is gv = 0, the weighting factor gv, which is p (x1, y1, z), gv = 1.0, and the weighting factor g associated with p (x1, y1, z + 1) is gv = 0.

Thus, the weighting factor gv can be determined directly from the extent of change in the CT value. In a method for directly determining the weighting factor gv, a large number of weighting factors gv have to be set for each amount of change of the CT value. Therefore, the amounts to be stored in the look-up table or the like increase in accordance with the amount of change of the CT value, so that the determination of the weighting factor gv becomes complex.

<Example of an index function>

7 and 8th FIG. 15 is graphs showing the index functions for determining in step S173 4 illustrate used indexes.

The index function after 7 (a) is a function in which the index changes linearly from 0 to 1 when the amount of change of the CT value is in a range between X1 and X3, and in which the index changes linearly from 1 to 0 when the amount of change of the CT Values in a range between X3 and X2. It is z. For example, assume that X1 equals 10 HU, X3 equals 90 HU, and X2 equals 170 HU. If p (x1, y1, z-1) = 10 HU, p (x1, y1, z) = 30 HU and p (x1, y1, z + 1) = 50 HU with respect to a particular image to be processed the amount of change of the CT value 40 HU. In this case, the index = 0.5 in the 7 (a) determined index function determined.

X1, X2 and X3 are set between 3 HU to 300 HU and 10 HU to 200 HU depending on an imaging condition. If they are 200 HU or more, this indicates a section or area that changes from a soft tissue to a bone or vice versa. If they are 10 HU or less, it means that the soft tissue continuously extends in multiple layer directions, or that the bone continuously extends in a plurality of layer directions. On the other hand, from the change of the CT value from 3 HU to 300 HU or the amount of change of the CT value from 10 HU to 200 HU, it is compulsory to generate the pinwheel type artifact or the cone beam artifact. Incidentally, the determination of the amount of change of the CT value may be appropriately changed based on the resolution, the film thickness or the table speed or the like during recording. When the amount of change of the CT value in the body axis direction moves in a range between 3 HU and 300 HU, it can be assumed as a result of various experiments that artifacts are generated.

The index function after 7 (b) is a function in which the index changes linearly from 0 to 1 when the amount of change of the CT value moves in a range between X1 and X3, then the index stays the same at 1 when the amount of change of the CT value is in a range between X3 and X4, and the index changes linearly from 1 to 0 as the amount of change of the CT value moves in a range between X4 and X2. It is z. B. assume that X1 equals 10 HU, X3 equals 40 HU, X4 equals 160 HU, and X2 equals 190 HU. According to the index function, in the case where the magnitude of the change of the CT value is in the range of 40 HU to 160 HU, it is judged that there is an artifact.

The index function after 7 (c) is a function in which the index changes from 0 to 1 according to a waveform, when the amount of change of the CT value falls within the range between X1 and X3, and then the index changes from 1 to 0 according to a waveform when the Change amount of the CT value falls within the range between X3 and X2.

On the other hand, what about the index function after 7 (d) As a result, the index is 1 when the amount of change of the CT value falls within the range between X1 and X2, and otherwise it is 0. Therefore, it means that an image to be processed is used as it is as a tomographic image when the amount of change of the CT value is X1 or less, or the amount of change of the CT value is X2 or more.

The index function after 8 (e) is a function in which, in the case where the magnitude of the change of the CT value falls within the range between 0 HU and X3, the index is 1 and in the case that the amount of change of the CT value is between X3 and X2 falls, the index changes linearly from 1 to 0. What the index function after 8 (f) If the magnitude of the CT value change falls between 0 HU and X2, and 0 otherwise, then the index is equal to 1. It is now assumed that X3 is in the range between 40 HU and 200 HU and X2 is in the range from 80 HU to 300 HU.

The statement that the magnitude of the change in the CT value is approximately 0 HU implies that the same soft tissue area or bone area continuously extends in multiple slice directions. Since the size of the CT value change is about 0 HU and the index = 1.0, it means that the associated pixel is averaged with a predetermined weighting factor according to equation (1). That is, in the range where the magnitude of the CT value change is between 0 HU and about 10 HU, noise reduction is performed without artifact reduction being performed. This means that the index function according to the 8 (e) or 8 (f) performs the artifact reduction using equation (1) in the range where the change in the CT value is in the range of over about 10 HU to under 200 HU or below 300 HU, and the noise reduction using the equation (1 ) is performed in the same manner in the case where the amount of change of the CT value is about 10 HU or less.

Although the index functions (a) to (f) in 7 and 8th As described above, not necessarily a single function needs to be used. It is possible to change the index function depending on the position in the z direction. For example, the index function (a) for a head region, the index function (c) for a neck region and the index function (d) for a leg region can be used. Even in the case of the flowchart used in 4 (b) is shown, the weighting factor gv can be set in a similar manner depending on the z-direction position.

<Specification of a tomographic image in which artifacts occur>

In the above embodiment, it has been estimated that when the amount of change of the CT value in the pixel p (x1, y1, z1) or in the pixel area (X1, Y1, Z1) in which plural pixels around the pixel p are combined in the predetermined range, artifacts are generated in this pixel, as in 5 is illustrated. The processing of equation (1) or (2) is accomplished on the pixel p (x1, y1, z1) to be processed. The following embodiment provides a method for further improving the specification or determination of an area where artifacts are generated.

9 FIG. 12 is a diagram illustrating tomographic images of the head of an object before undergoing an artifact reduction or reduction process and reconstruction areas P. FIG. In the upper area of tomographic image D3-A, as shown in FIG 9 (A1) is illustrated, many windmill artifacts are present, while the windmill artifacts in the lower region of the in 9 (B1) illustrated tomographic image D3-B almost no longer exist. The reconstruction area P, as in 9 is a 512x512 pixel square area that is parallel to an xy plane. As a result of the determination of the indices in the tomographic image D3-A and the tomographic image D3-B, pixels having an index = 1.0 are filled. In the reconstruction area P, as in 9 (A2) is illustrated, the artifact ratio obtained by dividing the number of Pixels with index = 1.0 is obtained by the total number of pixels, 0.12. In the reconstruction area P, as in 9 (B2) is illustrated, the artifact ratio obtained by dividing the number of pixels with index = 1.0 by the total number of pixels is 0.03. This calculation is performed by the artifact ratio calculation means 27 performed in 1 is illustrated.

If the processing of equation (1) or (2) is accomplished on each pixel of index = 1.0 regardless of the fact that the windmill-like artifact is almost absent in the tomographic image T3-B, the resolution in the z-axis is obtained. Direction reduced. Therefore, the artifact ratio calculating means checks 27 the artifact ratio corresponding to the ratio indicating how many pixels estimated to have the index = 1.0, that is, to have generated the artifacts, are picked up or occupied in the entire pixels in the reconstruction area P. When the artifact ratio is larger than a predetermined threshold SH, the determining means causes 29 the processing according to the equation (1) or (2) on the tomographic image D3. This means that the decision to recognize the artifact as generated is made under more stringent conditions.

10 11 is a diagram illustrating a flowchart for performing an artifact reduction process after artifact conditions. In the in 10 The flowcharts illustrated in FIG. 1 follow the same steps as those in the flowchart 4 (a) to reduce the artifacts assigned the same step names. An explanation of the steps with the same step name is partially avoided.

In step S171, the z position of an object HB that an operator wants to confirm is specified. The artifact reduction unit 25 determines a pixel p (x, y, z) to be processed.

In step S172, a change of the CT value in the z-direction at the pixel to be processed p (x, y, z) is measured.

In step S173, the index for the pixel to be processed p (x, y, z) is determined. When the reconstruction area P is a square area of 512 × 512 pixels, the amount of change of the CT value in the z-direction in a range where x is from 1 to 512 and a range where y is also from 1 to 512, measured.

Subsequently, the artifact reduction process proceeds to step S271 which takes a special step 9 represents. In step S271, the artifact ratio calculating means calculates 27 an artifact ratio. As for the artifact ratio, the ratio between pixels with index = 1 and all pixels (512 x 512) is calculated.

Incidentally, the area for the object HB is specified instead of all pixels, and the ratio can be calculated between the pixels with index = 1 and the number of pixels in this area. Instead of the ratio of the pixels with index = 1, the ratio of the points or digits with index = 0.7 or more or with index = 0.5 or more can be calculated. The following description is given in terms of the ratio of pixels with index = 1 with respect to all pixels (512 × 512) as the artifact ratio.

In step S272, the determining means determines 29 whether the artifact ratio is greater than the threshold SH. For example, the artifact ratio = 0.07 is used as the threshold SH. When the artifact ratio of a target tomographic image D3 is larger than the threshold value SH, the determining means proceeds to step S174. If this artifact ratio is smaller than the threshold value SH, the determining means goes to step S273.

In step S174, the pixel p (x, y, z) is processed on the basis of the index value according to the above equation (1) to determine a pixel p '(x, y, z) that follows from the processing. This calculation is performed on all the pixels (512x512) of the target tomographic image D3. In step S175, a tomographic image D31 (x, y, z) is obtained on the basis of the pixel p '(x, y, z) resulting from the artifact reduction process. Subsequently, the tomographic image D31 (x, y, z) is displayed 60 displayed.

On the other hand, the target tomography image D3 is displayed on the display at step S273 60 displayed while it is not subjected to image processing. This is because, although there is a likelihood that the pixels of index = 1 will cause the artifacts in the reconstruction area P, the artifacts are considered inconspicuous because the pixels of index = 1 in the entire reconstruction area P of FIG of low numbers, and because such image processing which reduces the resolution in the body axis direction is undesirable.

11 FIG. 12 is a cross-sectional view in a body axis direction from the chest of an object HB to its head in its upper portion and in its lower portion a graph showing the relationship between artifact ratios and the number of tomographic images arranged in the body axis direction.

When considering the relationship between the artifact ratios and the body axis direction, the artifact ratio of the breast to the environment of the eyes or eyebrows of the head (as indicated by a white dashed line in FIG 11 displayed) in the range of 0.9 to 2.2 or the like. In the tomographic image D3 (x, y, z), the artifact ratio between the surroundings of the eyes or eyebrows and the upper head end is in the range of 0.3 to 0.5 or so before the artifact reduction process according to the present embodiment is performed. How out 11 understandably, the more complex the shape of a structure, such as a bone, the more easily the artifacts. Conversely, if the shape of the bone or the like lying in the vicinity of the upper head end is simple, the artifacts hardly occur. In 11 the artifact ratio = 0.07 is defined as the threshold SH. Consequently, according to the flowchart 10 image processing is performed according to the equation (1) or the like on the tomographic image D3 in the vicinity of the breast to the vicinity of the eyes or eyebrows of the head. On the other hand, the image processing according to the equation (1) or the like is not performed on the part of the tomographic image D3 between the surroundings of the eyes or eyebrows and the upper head end, even if there are the pixels with index = 1.

Incidentally, the image reconstruction method according to the present embodiment may be a three-dimensional image reconstruction method based on the Feldkamp method as known today. Further, another three-dimensional image reconstruction method may be used. Alternatively, a two-dimensional image reconstruction method may be used. The image quality determined in each area varies according to the diagnostic applications, the preferences of an operator, etc. and is given in widely different forms. Therefore, the operator can set the settings of an imaging condition in advance in the most appropriate manner for each area.

Although the magnitude of the change in the CT value has been explained above using the difference between the maximum and minimum CT values of one or more layers in the vicinity of the pixel to be processed p (x1, y1, z1), it may be described below Using the average size of the CT value change obtained by dividing the difference between the maximum CT value and the minimum CT value by the number of layers.

In particular, the present embodiment is not limited to any particular form of scanning operation. This means that similar effects can be obtained even in the case of an axial scan, a cine-scan (cinematographic scan), a spiral scan, a variable pitch-factor spiral scan, and a helical shuttle scan. The present embodiment is not concerned with the inclination or inclination of the gantry 100 limited. This means that a similar effect can be achieved even in the case of a so-called inclined scan, in which the gantry 100 is inclined. The present embodiment can even be applied to the reconstruction of cardiac images which reconstructs each image in synchronism with a biological signal, in particular a cardiac signal.

Although the present embodiment is based on a medical X-ray CT apparatus 10 It may also be used for an X-ray CT-PET apparatus used in combination with an industrial X-ray CT apparatus or other apparatus, an X-ray CT-SPECT apparatus used in combination therewith , etc., are made available.

According to the present invention, an X-ray tomography device 10 which artificially filters out artifacts to reduce artifacts without reducing the directional resolution. The X-ray tomography device 10 contains a scanning device 101 . 103 for irradiating an object with X-rays, while at least one of a gantry 100 and / or a table 109 is moved along a body-axis direction of the object HB to thereby generate projection data of the object, a CT value change specifying means 26 for specifying the amount of change in the CT value in the body axis direction between a plurality of tomographic images obtained by backprojecting the projection data with respect to each of the tomographic images included pixel areas, an image processing condition selection means 28 for selecting an image processing condition for performing an image processing process for reducing artifacts according to the specified amount of change of the CT value and an artifact reduction means 25 for performing an image processing process using the image processing condition selection means 28 selected image processing condition.

LIST OF REFERENCE NUMBERS

Fig. 1

• 21 ... PREPROCESSOR, 23 ... BEAM CURING PROCESSOR, 24 ... THREE-DIMENSIONAL BACK PROJECTION PROCESSOR, 25 ... ARTEFAKTREDUKTIONSINHEIT, 26 ... CT CHANGE SPECIFICATION UNIT, 27 ... ARTEFAKTVERHÄLTNIS CALCULATION DEVICE, 28 ... IMAGE PROCESSING SELECTION DEVICE, 29 ... DESTINATION, 51 ... HIGH VOLTAGE GENERATOR, 53 ... SCANNER CONTROL DEVICE, 59 ... MEMORY DEVICE, 103 ... DETECTOR, 105 ... DATA TRANSFER DEVICE, 111 ... ROTATION MECHANISM, 113 ... GRINDING RING

Fig. 2

• CHANNEL DIRECTION
• FILM DIRECTION

Fig. 3

• BEGIN
• S11 ... DATA COLLECTION
• S12 ... PRE-PROCESSING PROCESS
• S13 ... BEAM CURING CORRECTION
• S14 ... EXECUTION OF THE Z-FILTER FOLDING PROCESS
• S15 ... EXECUTION OF THE RECONSTRUCTION FUNCTION FOLDING PROCESS
• S16 ... EXECUTION OF THE THREE-DIMENSIONAL BACK PROJECTION PROCESS
• S17 ... EXECUTION OF THE ARTEFACTREDUCTION PROCESS
• THE END

Fig. 4

(A)

• START OF THE ARTIFACT PROCESS
• S171 ... DETERMINATION OF THE PIXEL P (x, y, z) TO BE PROCESSED
• S172 ... MEASUREMENT OF THE SIZE OF THE CHANGE OF THE CT VALUE AT p (x, y, z - 1), p (x, y, z) AND p (x, y, z + 1)
• S173 ... CALCULATION OF THE INDEX VALUE ACCORDING TO THE CT CHANGE
• S174 ... PROCESSING OF THE PIXEL P (x, y, z) TO BE PROCESSED ACCORDING TO THE INDEX VALUE
• S175 ... DISPLAY OF THE TOMOGRAPHY IMAGE RESULTING FROM IMAGE PROCESSING
• THE END

(B)

• START OF THE ARTIFACT PROCESS
• S171 ... DETERMINATION OF THE PIXEL P (x, y, z) TO BE PROCESSED
• S172 ... MEASUREMENT OF THE SIZE OF THE CHANGE OF THE CT VALUE AT p (x, y, z - 1), p (x, y, z) AND p (x, y, z + 1)
• S174 ... PROCESSING OF THE PIXEL TO BE PROCESSED p (x, y, z) ACCORDING TO THE SIZE OF CT CHANGES
• S175 ... DISPLAY OF THE TOMOGRAPHY IMAGE RESULTING FROM IMAGE PROCESSING
• THE END

Fig. 7

• SIZE OF THE CHANGE IN THE CT VALUE, SIZE OF THE CHANGE OF THE CT VALUE, SIZE OF THE CHANGE IN THE CT VALUE, SIZE OF THE CHANGE IN THE CT VALUE,

Fig. 8

• SIZE OF CT CHANGES
• SIZE OF CT CHANGES

Fig. 9

• ARTIFACT RATIO

Fig. 10

• START OF THE ARTIFICIAL PRODUCTION PROCESS
• S171 ... DETERMINATION OF THE PIXEL P (x, y, z) TO BE PROCESSED
• S172 ... MEASUREMENT OF THE SIZE OF THE CHANGE OF THE CT VALUE AT p (x, y, z - 1), p (x, y, z) AND p (x, y, z + 1)
• S173 ... CALCULATION OF THE INDEX VALUE ACCORDING TO THE CT CHANGE
• S271 ... CALCULATING THE RELATIONSHIP BETWEEN PIXELS WITH INDEX = 1 (AS ARTEFAKT PIXELS) AND ALL PIXELS OR EFFECTIVE PIXELS IN ONE OBJECT
• S272 ... ARTIFACT RATIO> THRESHOLD SH
• S273 ... DISPLAY OF NON-PROCESSING TOMOGRAPHIC IMAGE D3
• S174 ... PICTURE PROCESSING ON THE PIXEL P (x, y, z) TO BE PROCESSED ACCORDING TO THE INDEX VALUE
• S175 ... DISPLAY OF THE TOMOGRAPHIC IMAGE D31 RESULTING FROM IMAGE PROCESSING
• THE END

Claims (10)

X-ray tomography device ( 10 ), comprising: a scanning device ( 102 . 103 ) for irradiating an object (HB) with X-rays, while at least one gantry ( 100 ) and / or a table ( 109 ) is moved along a body axis direction of the object (HB) to thereby generate projection data of the object (HB), a CT value change specifying means (FIG. 26 ) for specifying a magnitude of change in the CT value in the body axis direction between a plurality of tomographic images obtained by back projection of the projection data with respect to each of the pixel areas included in the tomographic images; an image processing condition selector ( 28 ) for selecting an image processing condition for performing an image processing process for reducing artifacts in accordance with the specified amount of CT value change; and artifact reduction means for performing an image processing process using the image processing condition selected by the image processing condition selection means. X-ray tomography device ( 10 ) according to claim 1, further comprising: an artifact ratio calculating means (10) 27 ) for calculating how each pixel area in which the size of the CT value change is in a predetermined range occupies a predetermined area included in each of the tomographic images, and determining means ( 29 ) for determining that if the ratio is greater than a threshold value, an image processing process for reducing the artifacts at the pixel area is performed. X-ray tomography device ( 10 ) according to claim 1, wherein the image processing process of the artifact reduction device ( 25 ) includes a process of multiplying the image areas of the tomographic images by weighting factors respectively selected based on the image processing condition and adding the results of the multiplications. X-ray tomography device ( 10 ) according to claim 3, wherein the image processing condition selection means (16) 28 ) sets the weighting factor greater than 0 when the size of the CT value change is in the predetermined range. X-ray tomography device ( 10 ) according to claim 3, wherein the image processing condition selection means (16) 28 ) changes the weighting factor according to the number of pixel areas in the body axis direction. X-ray tomography device ( 10 ) according to claim 3, wherein the image processing condition selection means (16) 28 ) changes the weighting factor according to the size of the change of the CT value. X-ray tomography device ( 10 ) according to any one of claims 1 to 6, wherein said CT value change specification means ( 26 ) determines an index value according to the size of the CT value change. X-ray tomography device ( 10 ) according to claim 7, wherein the plurality of pixel areas in the body axis direction are multiplied by weighting factors with respect to the pixel areas corresponding to the index values, and the results of the multiplications are added together. X-ray tomography device ( 10 ) according to claim 2, wherein said predetermined range is 3 HU to 300 HU. X-ray tomography device ( 10 ) according to claim 1, wherein each of the pixel areas comprises a single pixel or a plurality of pixels.

Images